text new page (beta)
English (pdf)
Article in xml format
Curriculum ScienTI
Automatic translation
Cited by SciELO 
Cited by Google 
Similars in
SciELO 
Similars in Google 
Permalink
Introduction
Benzodiazepines (BZD) comprise a subgroup of psychotropic drugs that act selectively to allosterically modulate gamma-aminobutyric acid subtype A (GABAA) receptor and mediate inhibitory synaptic transmission throughout the central nervous system1. They are commonly recommended for a variety of conditions such as anxiety, depression, somatic complaints, insomnia, alcohol withdrawal, delirium and violence and aggressive behavior in psychoses and disorders induced by neuroleptics2,3. The therapeutic indication for this group of drugs should be short term and for specific conditions such as those mentioned above.
Elderly people are more likely to use BZD4, but they feel less secure and have questionable clinical indications for taking BZD such as nonspecific emotional suffering5 or a chronic insomnia complaint. In several countries, prevalence studies demonstrate that chronic use of BZD in the elderly population is high, ranging from 3.9% to 35.9%6-8. This scenario has reached pandemic proportions for decades and is an important public health problem, because chronic use of this drug results in an increase in morbidity factors related to the risk of falls, intoxication and worsening of depressive symptoms and cognition9,10.
Depressive11,12 and anxiety disorders are frequent in the elderly, constituting an important source of emotional suffering and consequently the increased use of this pharmacological class8,13. Newer treatment consensus recommendations for depressive and anxiety disorders do not suggest BZDs as a first-line therapeutic14,15. The risk/benefit ratio increases when treating these disorders in the elderly, making the indication for BZD even more unfavorable. This is because of pharmacokinetic and pharmacodynamic changes that occur with aging, which may lead to an increased sensitivity of these individuals to the effects of BZD.
Thus, this study aims to examine the independent association between chronic BZD use in depression, anxiety and bipolar disorders, as well as other clinical and sociodemographic factors in a sample of elderly people who are enrolled in the Family Health Strategy (FHS).
Methods
Study design
This cross-sectional study was developed from the population-based survey entitled “The multidimensional study of the elderly in the family health strategy in Porto Alegre, Brazil (EMI-SUS)”16. The EMI-SUS was conducted from March, 2011 to December, 2012 and enrolled a random sample of elderly people who were participating in the FHS in Porto Alegre (RS/Brazil). Inclusion criteria were age ≥ 60 years and records registered in the FHS.
Data collection
The data collection procedure included identifying epidemiological and health data (sociodemographic, self-perception health, self-reported diseases, smoking, alcohol and pharmacotherapeutic evaluation) that were collected by community health agents at the homes of the elderly and during specialized psychiatric evaluation, which was carried out by professionals trained at the Hospital São Lucas of Pontifícia Universidade Católica do Rio Grande do Sul.
The mood disorder (major depression/dysthymia, bipolar) and anxiety diagnosis was made by psychiatrists using the DSM-IV criteria, and following the mental health evaluation protocol of the study17. The validated Brazilian version of the Mini-International Neuropsychiatric Interview (MINI) was used for evaluating psychiatric diagnoses18, and the psychometric properties of the instrument were considered satisfactory to excellent, with a good accuracy for anxiety and mood disorders in primary health care in Brazil19.
For pharmacotherapeutic evaluation, the participants were asked to specify all drugs used. In the interview conducted by the community health agent, this information was confirmed from prescriptions, drug packaging and medical records at the FHS. Drugs were coded according to the Anatomical Therapeutic Chemical (ATC) classification system recommended by the World Health Organization20. In this study, psychotropic medications included were BZD (N05BA, N03AE01), antidepressants (N06A, N06CA01), antiepileptic (N03A), antipsychotics (N05A) and other psychotropic drugs (N04AA02, N04BA01, N05BB01, N06BA07, N06BC01).
Sample size
The sample size of the study was calculated using a 0.05 significance level. Considering a target population of 22,000 elderly people enrolled by ESF in Porto Alegre, a minimum sample size of 491 elderly people was chosen, considering a 3.5% acceptable error for an expected prevalence of 20.0%.
Statistical analysis
Data were analyzed using Statistical Package for the Social Sciences (IBM SPSS Inc. Chicago, Illinois, version 17). The variables were described by the frequency, mean and standard deviation. Associations between categorical variables were tested using Pearson’s chi-square test. In specific cases, the chi-square test for linear tendency (ordinal variables with few categories) was used. To control for confounding variables and independence of variables, multivariate analysis was performed through Poisson regression.
Ethical considerations
This study was approved by the Ethical Research Committee of the Pontifícia Universidade Católica do Rio Grande do Sul (number 10/04967) and Porto Alegre Municipal Department of Health (registration 499/process 001.021434.10.7). All participants were informed of the objectives and research methods and they signed an informed consent form, according to the Guidelines and Norms Regulating Research of Resolution 196/96 of the National Health Council of the Ministry of Health.
Results
The 550 individuals included in the study were between 60 and 103 years of age (mean age, 68.6 ± 7.2 years), and comprised mostly females (63.1%). Most of these elderly people were married (37.8%), had incomplete primary education (69.1%), a little more than half of the individuals (55.0%) received less than one minimum wage (250 US dollars) and little more than half of the families (55.5%) received less than three minimum wages.
The prevalence of BZD use was 7.3%. This prevalence is compared with sociodemographic variables in Table 1. Those who had been widowed were found to use more BZD (10.8%) while single people used less BZD (1.1%; P = 0.044). There were no statistically significant differences in the other sociodemographic variables.
Table 1 Benzodiazepine (BZD) use compared with sociodemographic variables
| Sociodemographic variables | BZD use | P | |
|---|---|---|---|
| No n (%) | Yes n (%) | ||
| Gender | |||
| Female | 317 (91.4) | 30 (8.6) | 0.105† |
| Male | 193 (95.1) | 10 (4.9) | |
| Age (years) | |||
| 60-69 | 315 (92.6) | 25 (7.4) | 0.875‡ |
| 70-79 | 152 (92.7) | 12 (7.3) | |
| 80 or more | 43 (93.5) | 3 (6.5) | |
| Race | |||
| White | 320 (91.2) | 31 (8.8) | 0.270† |
| Black | 96 (97.0) | 3 (3.0) | |
| Brown | 71 (93.4) | 5 (6.6) | |
| Other | 15 (93.8) | 1 (6.3) | |
| Marital status | |||
| Married | 190 (92.7) | 15 (7.3) | 0.044† |
| Widowed | 141 (89.2) | 17 (10.8)2,1 | |
| Divorced | 83 (93.3) | 6 (6.7) | |
| Single | 90 (98.9) | 1 (1.1)-2,5 | |
| Education (years) | |||
| 0 | 79 (95.2) | 4 (4.8) | 0.299‡ |
| 1-7 | 341 (91.4) | 32 (8.6) | |
| 8 or more | 80 (95.2) | 4 (4.8) | |
| Individual income (minimum wage) | |||
| < 1 | 302 (91.8) | 27 (8.2) | 0.215‡ |
| 1 or more | 179 (94.7) | 10 (5.3) | |
| Total | 510 (92.7) | 40 (7.3) | |
† Pearson chi-square test; superscript numbers show results of residual analyses.
‡ Chi-square test for linear tendency.
Elders diagnosed with mood disorders represented 38.2% of the total population studied, with depression responsible for 28.8% and anxiety 20.2%. Elderly people without a diagnosis of mood disorder used less BZD (2.2%); however, those with depression (15.5%; P < 0.001) and anxiety (10.5%) used BZD more often. Those who self-identified and classified their health as poor/very poor used more BZD (21.2%; P = 0.003). Those who drank alcohol had a lower prevalence of use of BZD than those who did not drink alcohol (1.3%, P = 0.001; Table 2).
Table 2 Benzodiazepines(BZD) use compared with clinical and health variables
| Clinical and health variables | BZD | P | |
|---|---|---|---|
| No n (%) | Yes n (%) | ||
| Mood or Anxiety disorder | |||
| No | 307 (97.8) | 7 (2.2)-5,3 | < 0.001† |
| Depression | 87 (84.5) | 16 (15.5)3,6 | |
| Anxiety | 34 (89.5) | 4 (10.5) | |
| Depression and Anxiety | 45 (86.5) | 7 (13.5) | |
| Bipolarity | 27 (84.4) | 5 (15.6) | |
| Self-perceived health | |||
| Great/Good | 183 (94.8) | 10 (5.2) | 0.003‡ |
| Regular | 279 (93.6) | 19 (6.4) | |
| Poor/Very poor | 41 (78.8) | 11 (21.2)4,0 | |
| Smoker | |||
| No | 214 (93.4) | 15 (6.6) | 0.161† |
| Yes | 183 (93.8) | 12 (6.2) | |
| Ex-smoker | 98 (88.3) | 13 (11.7) | |
| Alcohol use | |||
| No | 327 (90.8) | 33 (9.2) | 0.001† |
| Yes | 150 (98.7) | 2 (1.3) | |
| Drug use | |||
| 0 | 74 (100.0) | 0 (0.0)-2,6 | < 0.001‡ |
| 1-2 | 111 (95.7) | 5 (4.3) | |
| 3-4 | 127 (93.4) | 9 (6.6) | |
| 5 or more | 196 (88.3) | 26 (11.7)3,3 | |
| Pharmacological classes | |||
| Antidepressants | |||
| No | 458 (96.8) | 15 (3.2) | < 0.001† |
| Yes | 52 (67.5) | 25 (32.5) | |
| Antipsychotics | |||
| No | 498 (93.6) | 34 (6.4) | < 0.001† |
| Yes | 12 (66.7) | 6 (33.3) | |
| Antiepileptics | |||
| No | 501 (93.1) | 37 (6.9) | 0.017† |
| Yes | 9 (75.0) | 3 (25.0) | |
| Others psychotropics | |||
| No | 494 (92.9) | 38 (7.1) | 0.524† |
| Yes | 16 (88.9) | 2 (11.1) | |
† Pearson chi-square test; superscript numbers show results of residual analyses.
‡ Chi-square test for linear tendency.
The average number of drugs used was 4.0 ± 2.9 (range, 0-13 drugs). Individuals who used 5 or more drugs showed a high prevalence of BZD use (11.7%, P < 0.001). Antidepressants (32.5%; P < 0.001), antipsychotics (33.3%; P < 0.001) and the antiepileptic (25.0%; P = 0.017) were the psychotropic classes that were most frequently used concomitantly with BZD (Table 2).
The final model of multivariate analysis was used to determine which variables were independently associated with the BZD use, and the results are presented in Table 3. Elderly people diagnosed with depression, anxiety, depression and anxiety concomitantly and bipolar disorder, and those who were using antidepressants had a higher risk of using BZD. Individuals who self-reported that they drank alcohol had a lower risk of BZD use.
Table 3 Final model of multivariate analysis using Poisson regression
| Variable | PR | CI 95% | P |
|---|---|---|---|
| Mood or Anxiety disorder | |||
| No | 1 | ||
| Depression | 2.92 | 1.08-7.85 | 0.034 |
| Anxiety | 7.06 | 2.44-20.44 | < 0.001 |
| Depression and Anxiety | 3.51 | 1.32-9.37 | 0.012 |
| Bipolarity | 3.54 | 1.04-12.11 | 0.044 |
| Antidepressants | |||
| No | 1 | ||
| Yes | 8.60 | 4.14-17.89 | < 0.001 |
| Alcohol use | |||
| No | 1 | ||
| Yes | 0.23 | 0.06-0.94 | 0.040 |
PR: prevalence ratio; CI: confidence interval.
Discussion
Large-scale BZD use has been widely accepted worldwide, because these drugs have been considered to be effective as anxiolytics and they are safer than the drugs that were previously available, such as barbiturates. The benefit of a lower toxicity and less potential to develop a chemical dependency contributed to the widespread BZD use over the past decades; this transformed a “benefit” into an important public health problem, especially in the elderly who are typically the main consumers this type of drug.
The prevalence of BZD use (7.3%) is considered high. Brunoni et al. presented data from six universities located in different Brazilian regions (São Paulo, Rio de Janeiro, Salvador, Porto Alegre, Belo Horizonte and Vitória), where they detected a BZD use prevalence of 3.9% (in those 35 to 75 years of age), and older people were the most likely to use BZD (OR 3.48)8. The prevalence was even higher (21.7%) in an elderly community sample of residents of the city of Bambuí, Minas Gerais, Brazil21. Prevalence rates in other countries ranged from 16% in Australia22 to 31% in Finland7 and 36% in Canada6. These results are particularly important because there are guidelines that classify the BZD use as inappropriate, particularly because of side effects in the elderly23.
The main socio-demographic characteristics related to chronic BZD use were female (8.6%), white (8.6%), widowed (10.8%) and income less than the one minimum wage (8.2%). Only the marital status showed a significant difference, with those who had been widowed using more BZDs. In this group, the majority were women who had depression and used antidepressants, which are factors that are strongly related to BZD use24. Previous studies that examined the role of living alone or marital status obtained similar results13,25. This finding was not an independent factor that was associated with outcome in a multivariate analysis.
The highest frequency of BZD use is associated with a diagnosis of mood or anxiety disorders, poor self-perceived health, the use of five or more drugs and concomitant use of other psychotropic drugs, particularly antidepressants, antipsychotics and antiepileptics. In a multivariate analysis, the factors that remained were diagnosed with depression (RP: 2.92), anxiety (RP: 7.06), depression and anxiety (RP: 3.51) and bipolar disorder (RP: 3.54). When considering these diagnoses, the elderly person with these psychiatric disorders can be considered very likely to receive BZD. These drugs are indicated for use in various syndromes that present as nonspecific emotional suffering. However, in almost all situations, the BZD use is contraindicated in elderly people23, as well as in the general population. BZDs are not indicated for moderate to severe depression, and there is also no evidence to support their use in minor depression26. Toxic effects, cognitive dysfunction, risk of worsening depression and fall hazards, among others, will probably outweigh any positive effect of the BZD in the elderly.
According to some authors, the concomitant BZD and antidepressant use can be considered a strategy to increase treatment effectiveness27. Short-term use is the most widely accepted use of BZD to treat depression and anxiety, mainly to achieve rapid relief of symptoms at the start of therapy. There is subsequent reduction of the antidepressant dose when it starts to show its effect, thus improving the adherence to antidepressant therapy28. The rationale for combination treatment is multidimensional, including neurological bases and clinical factors, because of different pharmacokinetic mechanisms and clinical effects. However, this must be balanced because there is the potential for BZD dependency, and there are antidepressants with anxiolytic/sedative effects that can be used instead of BZDs. This association also raises the issue of polypharmacy, which should be an exception in treating elderly people. In multivariate analysis, antidepressants were most often associated with BZD, showing a high chance of co-prescription (RP: 8.60). Data from a Dutch cohort study of people with depression and anxiety found a milder association between BZD and antidepressants (OR: 3.5)13, while another Brazilian study showed a strong association (OR: 7.95)8.
BZDs have been reported to be associated with alcohol consumption24. This combination is important because of the possibility of mood disorders, excessive sedation, increased risk of falls, memory problems and traffic accidents, especially in the elderly. In contrast, our results showed that older people who used BZD consumed less alcohol. These elderly people potentially should have been instructed not to consume alcohol during treatment.
This study is subject to some limitations, as follows: 1) the cross-sectional design is limited to establishing cause and consequence; 2) while the diagnostic examination is important, overestimation of the diagnosis is recognized in structured interviews based on current diagnostic systems29; and 3) the results of the multivariate analysis should be interpreted with caution because of the small absolute number of individuals with diagnoses who were investigated.
In conclusion, BZDs are often used by elderly people over long periods of time. Elderly people who make take multiple drugs, especially antidepressants, are more likely to use BZDs, as are those with a clinical diagnosis of depression, anxiety, depression/anxiety and bipolar disorders. These questions have a direct impact on an increase in morbidity that results from negative effects of psychotropic drug over-prescription and mistreatment of mood and anxiety disorders.
